Classification of vasovagal syncope from physiological signals on tilt table testing

Mahbuba Ferdowsi,Ban-Hoe Kwan,Maw Pin Tan,Nor’ Izzati Saedon,Sukanya Subramaniam,Noor Fatin Izzati Abu Hashim,Siti Sakinah Mohd Nasir,Imran Zainal Abidin,Kok Han Chee,Choon-Hian Goh
DOI: https://doi.org/10.1186/s12938-024-01229-9
2024-04-01
BioMedical Engineering OnLine
Abstract:The diagnostic test for vasovagal syncope (VVS), the most common cause of syncope is head-up tilt test (HUTT) assessment. During the test, subjects experienced clinical symptoms such as nausea, sweating, pallor, the feeling of palpitations, being on the verge of passing out, and fainting. The study's goal is to develop an algorithm to classify VVS patients based on physiological signals blood pressure (BP) and electrocardiography (ECG) obtained from the HUTT.
engineering, biomedical
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to develop an algorithm for classifying patients with vasovagal syncope (VVS) based on physiological signals (such as blood pressure and electrocardiogram) obtained during the head - up tilt test (HUTT). Specifically, the research objectives are: 1. **Identify key physiological indicators**: The study hypothesizes that systolic blood pressure (SBP), diastolic blood pressure (DBP) and their time - domain and frequency - domain variables may be more important predictors than heart rate (HR) and heart rate variability (HRV) for detecting VVS. 2. **Improve classification accuracy**: By combining multiple data processing techniques (such as data imputation and feature selection) and using six machine - learning classifiers (support vector machine SVM, K - nearest neighbor KNN, Gaussian naive Bayes GNB, multinomial naive Bayes MNB, logistic regression LR and random forest RF), improve the accuracy of VVS classification. 3. **Enhance model interpretability**: Utilize explainable artificial intelligence (XAI) techniques, especially partial dependence plot (PDP), to enhance the interpretability and transparency of the model. Through these methods, the study aims to develop an algorithm that can effectively classify VVS patients, thereby improving the diagnosis and treatment process of VVS and bringing practical benefits to healthcare providers and patients.